Part 1: BMJ Investigation reveals serious flaws in trials of blockbuster anti-clotting drug
The anti-clotting drug ticagrelor was approved on flawed studies—and remains widely used today despite evidence of scientific misconduct and regulatory failure.
For more than a decade, millions of people around the world have taken TICAGRELOR—an anti-clotting drug to reduce the risk of heart attack or stroke.
It’s sold as Brilinta in the US and Australia, and Brilique in Europe.
Doctors routinely prescribe it for people after stenting, heart attacks or acute coronary syndrome—a condition where the blood flow to the heart is blocked.
Marketed as faster and more effective than its cheaper competitor clopidogrel (Plavix), ticagrelor became a multibillion-dollar success for AstraZeneca and was rapidly adopted into cardiology guidelines across the globe.
But that foundation is now collapsing.
A series of investigations by senior editor Peter Doshi at The BMJ has uncovered serious irregularities in both the pivotal trial that launched the drug and two key studies that underpinned its mechanism of action.
The new evidence raises urgent questions about whether ticagrelor should ever have been approved—and why it remains on the market today.
The trial that launched a blockbuster
In 2024, Doshi published a forensic analysis of the PLATO trial—the pivotal study used to win approval for ticagrelor.
At first glance, the results seemed promising: ticagrelor reduced cardiovascular deaths compared to clopidogrel. But when the data were broken down by region, the story began to unravel.
PLATO was a global trial conducted across 43 countries.
In the US—where trial sites were monitored by independent contractors—patients on ticagrelor actually fared worse. FDA records showed a 27% higher risk of major cardiac events compared to those on clopidogrel.
By contrast, in European countries like Hungary and Poland, the results strongly favoured ticagrelor.
What accounted for the discrepancy?
Unlike the independently monitored US sites, the European trial centres were overseen by AstraZeneca itself.
That the drug performed better at sites managed by its sponsor raises serious concerns about site conduct and the reliability of the trial data.
Further problems emerged in how the trial outcomes were adjudicated.
A special committee was tasked with determining whether deaths and heart attacks qualified as major events. Doshi found that 45 heart attacks were added to the clopidogrel group during adjudication—while none were added to the ticagrelor group.
Of 20 deaths where adjudicators could not agree on the cause, 17 occurred in the clopidogrel arm. In six of those cases, the final classification was changed in a way that favoured ticagrelor—without explanation.
The pattern of one-sided changes raised a troubling question: was the trial truly blinded?
The adjudication committee was supposed to be unaware of treatment allocation, but the bias in favour of ticagrelor suggests that blinding may have been compromised.
The trial’s senior investigators refused to comment. AstraZeneca, which funded the trial, managed the data and employed several of the study’s authors, said it had “nothing to add.”
Some within the FDA also voiced alarm.
Internal records show that medical officer Dr Thomas Marciniak strongly advised against approving ticagrelor, calling AstraZeneca’s submission “one of the worst he’d ever seen.”
He noted that “ticagrelor patients undergoing early percutaneous coronary intervention… fare[d] worse than clopidogrel patients.”
But his concerns were ignored.
Despite glaring red flags—regional inconsistencies, questionable adjudication, and internal dissent at the FDA—the trial sailed through regulatory review.
Ticagrelor was granted approval in 2011 and unleashed onto the global market.
Uncovering misconduct in key platelet studies
In a follow-up investigation published last month, Doshi turned his attention to two key studies—ONSET/OFFSET and RESPOND.
These “platelet function” studies weren’t designed to assess clinical outcomes, but were essential in marketing the drug as a faster-acting, more potent alternative to clopidogrel.
But these studies were also deeply concerning.
In the RESPOND study, for example, Doshi found that researchers reported a different primary endpoint from the one prespecified in the protocol—without disclosing the change.
The original primary outcome showed no significant benefit for ticagrelor (P=0.157), but the published paper instead reported a more favourable result (P=0.005).
Lead author of this study, Professor Richard Storey, would not respond to my inquiries.
Doshi asked investigators from the ONSET/OFFSET and RESPOND studies to share their original platelet data, but the three he contacted all said they no longer had the data—claiming it had been archived or destroyed.
However, The BMJ accessed readouts from some lab machines that had stored platelet test results. Analysis revealed that more than 60 of the 282 readings from one site were missing from the dataset submitted to the FDA.
They also found striking anomalies—for example, some participants showed increased platelet activity after taking ticagrelor, a paradoxical result for an anti-clotting drug.
According to Doshi, the investigators performed an "unpublished data adjustment" which rendered the irregularities invisible.
Dan Atar, editor-in-chief of Cardiology, told The BMJ that such an adjustment “absolutely should have been reported,” adding that without transparency, “one cannot even evaluate its appropriateness.”
Who really ran the trials?
The deeper Doshi investigated, the more the studies unravelled.
One person listed as a co-author of the RESPOND study told The BMJ, “I did not participate in the RESPOND study.” Another investigator who enrolled 12 patients wasn’t listed as an author at all.
Most investigators, including the principal investigator, either declined to speak or were unreachable.
Dr Paul Gurbel, the lead investigator on both studies, declined repeated requests for interview. So did the head of his hospital’s institutional review board.
A spokesperson for Gurbel’s hospital told The BMJ that “any allegations of any research misconduct… are baseless and erroneous.”
Both the ONSET/OFFSET and RESPOND studies were published in Circulation, one of cardiology’s leading journals.
Lack of transparency
Victor Serebruany, a pharmacologist who once collaborated with AstraZeneca, now speaks out against the drug.
He told The BMJ there are episodes of “skyrocketing rebound and profound platelet inhibition” with ticagrelor, making patients prone to bleeding or blood clots.
“If doctors had known what happened in these trials, they would never have started using ticagrelor,” he added.
Serebruany described the FDA’s failure to act on these warning signs as “unconscionable,” adding “we all need to know how and why that happened.”
How is this drug still on the market?
Doshi’s investigations have painted a damning picture.
The PLATO trial that underpinned ticagrelor’s approval is riddled with anomalies.
And the platelet studies that reinforced its mechanism of action contain missing data, undisclosed outcome changes, and authorship inconsistencies.
One FDA medical reviewer warned that the data were not trustworthy—yet the agency approved the drug anyway.
The result? Ticagrelor remains in widespread use.
Now, with generic versions entering the market, ticagrelor’s reach will only grow. But its scientific foundation appears irreparably compromised.
When companies can rewrite outcomes, suppress data, and still win regulatory approval, what’s the point of calling these trials the ‘gold standard’?
If ticagrelor doesn’t meet the threshold for withdrawal, what drug ever will?
See Part 2: Journal ignores evidence of misconduct in ticagrelor studies [FULL STORY]
[Unlike the independently monitored US sites, the European trial centres were overseen by AstraZeneca itself.]
always good when you get to mark your own homework !
Monstrous. And dare I say business as usual?